GWAS Bioinformatics MCP Server

What GWAS Bioinformatics MCP Server MCP server does

GWAS Bioinformatics MCP Server MCP server gives MCP-compatible AI clients access to biological information from 14 named databases. Its tools cover protein and gene records, clinical variants, genome-wide association studies, expression QTL data, protein interactions, domains, structures, pathways, drug relationships, pharmacogenomics, and genetic diseases.

The server can search UniProt, NCBI Gene, and Ensembl; retrieve ClinVar interpretations and variant information; query the GWAS Catalog and GTEx; inspect STRING interactions and InterPro domains; search AlphaFold and PDB structures; retrieve KEGG pathways; and investigate Open Targets, PharmGKB, and OMIM records.

How it works

The package runs as a Python MCP server and exposes database operations as named tools. An MCP client sends a request such as a gene search, rsID lookup, pathway query, or structure search, and the server calls the relevant external database service. The available tools are grouped around proteins and genes, clinical variants, interactions, structures and pathways, drug discovery, and genetic diseases.

API responses are cached for one hour. Calls are asynchronous and non-blocking, and the server includes handling for API failures.

Setup and configuration

Install the published package with:

pip install gwas-mcp

A Claude Desktop configuration can start the server with the Python module entry point:

{
  "mcpServers": {
    "gwas-bioinformatics": {
      "command": "python",
      "args": ["-m", "gwas_mcp.server"]
    }
  }
}

After saving the configuration, restart Claude Desktop. The documented configuration paths are under the Claude application data directory on Windows, macOS, and Linux.

For development, clone the repository, install it in editable mode with pip install -e ., and start it with python -m gwas_mcp.server.

Tools and capabilities

Representative tools include:

• search_uniprot, get_protein_details, search_ncbi_gene, and search_ensembl_gene for gene and protein research.
• get_variant_info, search_clinvar, get_clinvar_variant, and annotate_snps for variant information and functional consequences.
• query_gwas_catalog and get_eqtl_data for association and expression-QTL queries.
• get_protein_interactions, get_interaction_network, and get_functional_enrichment for interaction and enrichment analysis.
• get_alphafold_structure, search_pdb_structures, and pathway tools for structural and biological context.
• get_drug_targets, get_disease_associations, search_pharmgkb, and get_drug_gene_interactions for drug and disease research.
• search_omim and get_gene_diseases for genetic disease lookups.

GWAS Bioinformatics MCP Server MCP server is therefore suited to exploratory analysis across multiple sources, such as checking a gene, tracing associated variants, reviewing pathways, and finding related drugs or diseases in one client session.

Limitations and notes

The README identifies the supported databases and tools but does not document authentication requirements, rate limits, database version pinning, result pagination, or the freshness of each source. Results depend on the external services called by the tools. The project requires Python 3.10 or newer according to its documentation, and it is licensed under MIT.